J. Phys. Chem. B 2010, 114, 12995–13004
12995
Effect of Watson-Crick and Hoogsteen Base Pairing on the Conformational Stability of C8-Phenoxyl-2′-deoxyguanosine Adducts Andrea L. Millen,† Cassandra D. M. Churchill,† Richard A. Manderville,‡ and Stacey D. Wetmore*,† Department of Chemistry, UniVersity of Lethbridge, 4401 UniVersity DriVe, Lethbridge, Alberta, Canada, T1K 3M4, and Department of Chemistry, UniVersity of Guelph, Guelph, Ontario, Canada, N1G 2W1 ReceiVed: June 23, 2010; ReVised Manuscript ReceiVed: August 16, 2010
Bulky DNA addition products (adducts) formed through attack at the C8 site of guanine can adopt the syn orientation about the glycosidic bond due to changes in conformational stability or hydrogen-bonding preferences directly arising from the bulky group. Indeed, the bulky substituent may improve the stability of (non-native) Hoogsteen pairs. Therefore, such adducts often result in mutations upon DNA replication. This work examines the hydrogen-bonded pairs between the Watson-Crick and Hoogsteen faces of the ortho or para C8-phenoxyl-2′-deoxyguanosine adduct and each natural (undamaged) nucleobase with the goal to clarify the conformational preference of this type of damage, as well as provide insight into the likelihood of subsequent mutation events. B3LYP/6-311+G(2df,p)//B3LYP/6-31G(d) hydrogen-bond strengths were determined using both nucleobase and nucleoside models for adduct pairs, as well as the corresponding complexes involving natural 2′-deoxyguanosine. In addition to the magnitude of the binding strengths, the R(C1′ · · · C1′) distances and ∠(N9C1′C1′) angles, as well as the degree of propeller-twist and buckle distortions, were carefully compared to the values observed in natural DNA strands. Due to structural changes in the adduct monomer upon inclusion of the sugar moiety, the monomer deformation energy significantly affects the relative hydrogenbond strengths calculated with the nucleobase and nucleoside models. Therefore, we recommend the use of at least a nucleoside model to accurately evaluate hydrogen-bond strengths of base pairs involving flexible, bulky nucleobase adducts. Our results also emphasize the importance of considering both the magnitude of the hydrogen-bond strength and the structure of the base pair when predicting the preferential binding patterns of nucleobases. Using our best models, we conclude that the Watson-Crick face of the ortho phenoxyl adduct forms significantly more stable complexes than the Hoogsteen face, which implies that the anti orientation of the damaged base will be favored by hydrogen bonding in DNA helices. Additionally, regardless of the hydrogen-bonding face involved, cytosine forms the most stable base pair with the ortho adduct, which implies that misincorporation due to this type of damage is unlikely. Similarly, cytosine is the preferred binding partner for the Watson-Crick face of the para adduct. However, Hoogsteen interactions with the para adduct are stronger than those with natural 2′-deoxyguanosine or the ortho adduct, and this form of damage binds with nearly equal stability to both cytosine and guanine in the Hoogsteen orientation. Therefore, the para adduct may adopt multiple orientations in DNA helices and potentially cause mutations by forming pairs with different natural bases. Models of oligonucleotide duplexes must be used in future work to further evaluate other factors (stacking, major groove contacts) that may influence the conformation and binding preference of these adducts in DNA helices. Introduction Natural and exogenous sources of DNA damage result in a variety of DNA modifications, the most common including nucleobase oxidation,1 alkylation,2 and deamination.3,4 Bulky substituents derived from many environmentally prevalent compounds also frequently add to the natural DNA nucleobases to generate so-called DNA adducts (addition products).5 Guanine adducts formed through modifications to the N2, N7, O6, and C8 sites (Figure 1a) are common, where aryl adducts produced via attachment to C8 are currently of particular interest in the literature.6-14 These adducts can arise from carcinogenic aryl hydrazines6,15 and polycyclic aromatic hydrocarbons (PAHs),16 which are metabolized to produce aryl radical intermediates that frequently attack guanine. Additionally, some natural phenolic † ‡
University of Lethbridge. University of Guelph.
compounds, such as the possible human carcinogen ochratoxin A, can be activated to form C8-bonded adducts.17-22 Our group is currently interested in C8-aryl adducts formed due to phenolic toxins23-27 that are metabolically activated to unstable phenoxyl radical intermediates, which directly react with guanine to form both covalent O- and C-bonded adducts.28 Since formation of DNA adducts is often the initiation event in the carcinogenic process, it has been proposed that C8 deoxyguanosine adducts are major contributors to the toxic effects of phenolic compounds.28 In this light, our most recent work has focused on the structure and properties of the ortho and para C8-bonded phenoxyl-2′-deoxyguanosine adducts (Figure 1b and c).25,29,30 Bulky DNA adducts exert their biological effects in a variety of ways. However, the preferred orientation of the modified base about the glycosidic bond often contributes to their mutagenicity. In particular, depending on the nature and location of the bulky substituent, the preferred anti orientation (Figure 1, left) of the
10.1021/jp105817p 2010 American Chemical Society Published on Web 09/20/2010
12996
J. Phys. Chem. B, Vol. 114, No. 40, 2010
Figure 1. The structure and chemical numbering of (a) natural 2′deoxyguanosine and the (b) ortho and (c) para C8-phenoxyl-2′deoxyguanosine adducts considered in the present work. The dihedral angle χ (∠(O4′C1′N9C4), purple) defines the glycosidic bond orientation to be anti (left, Watson-Crick bonding faces in blue) when χ ) 180 ( 90° or syn (right, Hoogsteen bonding faces in red) when χ ) 0 ( 90°. θ (∠(N9C8C10C11), green) defines the degree of twist between the nucleobase and the bulky substituent (θ ) 0 or 180° represents a planar structure).
natural base may become destabilized relative to the syn orientation (Figure 1, right), which subsequently leads to mispairing and possibly mutations.10,31-38 For this reason, we previously studied the conformational preferences of the ortho and para phenoxyl adducts in a variety of (nucleobase,25 nucleoside,25,29,30 and nucleoside 5′-monophosphate30) model systems. Results from our most accurate models imply that the energy difference between the anti and syn conformations for both adducts is likely quite small in DNA helices.30 We therefore proposed that both orientations may be present in DNA, and that environmental factors (sequence effects, stacking and hydrogen-bonding interactions, etc.) may have a large influence on the population of a given conformer. The effect of hydrogen-bonding interactions on the anti/syn conformational stability of bulky nucleobase adducts in DNA helices is particularly interesting, since the preference for a specific hydrogen-bonded pair and the conformation of the adduct are mutually dependent. Specifically, the Watson-Crick face is available for hydrogen bonding to the neighboring strand in the helix when the adduct adopts the anti orientation (Figure 1, left), while only the Hoogsteen face of the adduct is accessible for hydrogen-bond pair formation when the adduct adopts the syn orientation (Figure 1, right). Furthermore, it has been experimentally observed that the addition of a bulky substituent to C8 of the purines enhances Hoogsteen interactions due to the presence of additional hydrogen-bonding sites.39,40 Therefore, bulky C8 adducts could have preferred hydrogen-bonding patterns that are inaccessible to natural DNA, which may change the structural preference of the base. Specifically, if one conformation of the bulky adduct forms a relatively strong hydrogen-bonded complex with a natural base, then this particular base pair may be preferentially incorporated into the DNA helix upon replication and the adduct may preferentially adopt the corresponding conformation.
Millen et al. Evidence for the close relationship between the strength of the hydrogen-bonding interactions and the preferred nucleotide conformation in DNA exists in the literature. Lesion bypass polymerases have been found to allow either the anti or syn orientation of template bases in the active site, and interstrand hydrogen-bonding interactions determine the nature of the incoming nucleotide.41,42 Although strong guanine-cytosine (G-C) hydrogen bonds normally anchor guanine in the anti orientation, simulations show that G-G pairs between one anti G and one syn G are frequently misincorporated (by polymerase χ from African swine fever virus) due in part to their strong Hoogsteen bonding, as well as lack of distortion, compared to other mispairs.43 Since modifications to the guanine base often induce the syn orientation,35,37,38,44-48 DNA damage may influence the base-pairing preference. For example, 8-oxo-2′deoxyguanosine, a frequent product of oxidative damage,1 is known to adopt a syn orientation that forms strong Hoogsteen bonds to adenine and therefore results in G-C to T-A (thymine-adenine) mutations.44,49 In addition, simulations of a flexible guanine lesion (5-guanidino-4-nitroimidazole) in a DNA helix show that the nature of the hydrogen-bonded pair can determine the preference for the anti or syn conformation.50 Therefore, in cases where there is a small energy difference between nucleotide conformations, the relative strengths of possible hydrogen-bonded complexes can provide insight into the preferred conformation and the associated mutagenic potential of the damaged base. Interestingly, subtle structural changes between bulky nucleobase adducts can lead to very different mutational profiles due to unique conformational equilibria and hydrogen-bonding patterns.45,51,52 For example, the carcinogenic N-2-(2′-deoxyguanosine-8-yl)-acetylaminofluorene (G-AAF) lesion is known to exhibit frameshift mutations due to the syn orientation, where the bulky group intercalates into the helix and displaces cytosine.33,46 However, the closely related N-(2′-deoxyguanosine8-yl)-aminofluorene (G-AF) adduct, which differs only by an acetyl group, is believed to readily adopt both the syn and anti conformations.33,36,46,47 Although the anti orientation of G-AF maintains the Watson-Crick hydrogen-bonding pattern of guanine and cytosine, the syn conformation is believed to lead to point mutations due to a Hoogsteen bonding arrangement with adenine.53,54 Despite the similar bulkiness of these adducts, the major conformation of the C8-aminobiphenyl adduct is the anti orientation, while the syn orientation is infrequently adopted.55 Therefore, it is important to individually assess the base pairing of each bulky adduct in all possible conformations. The present work considers the Watson-Crick and Hoogsteen hydrogen-bond strengths between the ortho or para C8-phenoxyl guanine adduct and the four natural DNA bases. As discussed above, it is important to consider both the ortho and para adducts, since their rather subtle structural differences may lead to changes in the (Hoogsteen) hydrogen-bonding preference. Therefore, each adduct may adopt different conformations and exhibit unique mutagenic properties. In the absence of experimental data regarding the base-pairing preferences of these adducts, our results will allow us to determine the potential role of base pairing in dictating the most favorable conformation of each adduct in DNA helices. Additionally, since previous studies have successfully used calculated binding strengths of damaged bases to their natural counterparts to predict the most probable DNA mutations,56-59 our work represents the first step in modeling the potential mutagenic profile of phenoxyl adducts, as well as highlighting systems of interest for further study in a duplex environment.
Stability of C8-Phenoxyl-2′-deoxyguanosine Adducts
J. Phys. Chem. B, Vol. 114, No. 40, 2010 12997
Computational Details The Watson-Crick and Hoogsteen hydrogen-bonded complexes formed between the anti or syn conformations of the ortho or para C8-phenoxyl-2′-deoxyguanosine adducts and the natural bases were optimized in the gas phase with B3LYP/ 6-31G(d). This level of theory has been found to provide accurate geometries for hydrogen-bonded complexes involving the natural nuclebases.60 Two computational models were used for the adduct base pairs: (1) the nucleobase model, where the DNA sugar moiety was replaced with a hydrogen atom for both the adduct and the natural bases, and (2) the nucleoside model, where a deoxyribose unit was attached to the damaged base and a methyl group replaced the sugar moiety in the natural nucleoside. In the adduct nucleoside, the ∠(C4′C5′OH) dihedral angle was constrained to 180° to prevent intramolecular interactions between the C5′-hydroxyl group and the base that are non-native to DNA, where the C5′ hydroxyl is a phosphate group. When the natural nucleoside is paired with the adduct, the sugar attached to the natural base is reduced to a methyl group since the structures of the natural bases are not greatly affected by the presence of the sugar.61 Inclusion of both the sugar and the methyl group allows the R(C1′ · · · C1′) distance, which dictates the width of the helix, and the ∠(N9C1′C1′) angle, which dictates the base-pair opening angle with respect to N9 in the adduct, to be measured and compared to average experimental values for B-DNA double helices (10.6 Å and 55°, respectively).62 Although the final optimized geometries reported in this work may deviate from those expected for B-DNA, all initial guesses were built to have R(C1′ · · · C1′) distances and ∠(N9C1′C1′) angles appropriate for B-DNA. In some cases, this involved multiple starting orientations, where the resulting lowest energy structure was selected for detailed analysis in this study. For comparison, identical calculations were performed using the nucleoside model, where the adduct was replaced with natural 2′-deoxyguanosine. Only the nucleoside model was used for the natural guanine complexes, since this model best allows for a structural comparison of natural and damaged hydrogen-bonded complexes (see the Results and Discussion for further details). B3LYP/6-311+G(2df,p) singlepoint calculations were performed to obtain accurate binding energies, which also include counterpoise corrections to account for the basis set superposition error (BSSE).63 In all cases, the magnitude of the BSSE correction was less than 3 kJ mol-1. The interaction energies were calculated as the difference between the energy of the optimized complex and the sum of the energies of the monomers in the geometry adopted in the complex (EInt). Since the monomers in this orientation are not fully optimized, the resulting binding energies (EInt) do not include zero-point vibrational energy (ZPVE) corrections. Subsequently, the deformation energy (EDef) of the monomers upon complexation was included in the hydrogen-bond strengths, such that ETot ) EInt + EDef. In other words, the total interaction energies (ETot) were calculated as the difference between the energy of the optimized complex and the sum of the energies of the individually optimized monomers. To allow more direct comparisons of EInt and ETot, the binding energies calculated using the individually optimized monomers (ETot) discussed in the main text do not include ZPVE corrections. Nevertheless, the values of ETot that include scaled (0.9806)64 ZPVE corrections are provided in the data tables (in parentheses) for comparison, where the corrections range between 2.3 and 6.3 kJ mol-1. All calculations were performed using Gaussian 03.65
Figure 2. B3LYP/6-31G(d) structures (distances in Å, angles in deg) for complexes between the (a) Watson-Crick (anti) and (b) Hoogsteen (syn) faces of natural 2′-deoxyguanosine and each of the four natural DNA nucleoside models.
Results and Discussion 1. Natural 2′-Deoxyguanosine. A great deal of computational work has considered the strengths and binding arrangements between the natural nucleobases.59,61,66-92 However, these studies used a variety of bases, base-pairing modes, computational models, and levels of theory. Therefore, before discussing the hydrogen-bonding patterns of the modified bases, it is necessary to systematically examine the structure and stability of the particular hydrogen-bonded complexes between 2′deoxyguanosine and each of the four natural (N1- or N9methylated) DNA nucleosides of interest in this work (Figure 2). This information will subsequently be used to illustrate the differences in behavior between natural and damaged guanine. For example, it is important to consider whether the interaction energies of the natural pairs are weakened or strengthened by the bulky group. Additionally, this data will help us determine whether inclusion of the bulky group permits new hydrogenbonding patterns. For example, Hoogsteen complexes between the natural bases generally do not meet the structural criteria for an optimal fit in the DNA helix,42 but interactions with bulky substituents may significantly alter the geometries of these base pairs and thereby produce more viable structures. First, the hydrogen-bond strengths of base pairs between the Watson-Crick face (anti orientation about the glycosidic bond (Figure 1a, left)) of 2′-deoxyguanosine and the neighboring DNA strand will be discussed. The natural 2′-deoxyguanosinecytosine (dG-C) Watson-Crick pair contains three strong hydrogen bonds (Figure 2a) and has a calculated (EInt) interaction strength of -113 kJ mol-1 (Table 1), which is in good agreement with previous literature.67-69,74 In contrast, the 2′-deoxyguanosine
12998
J. Phys. Chem. B, Vol. 114, No. 40, 2010
Millen et al.
TABLE 1: B3LYP/6-311+G(2df,p)//B3LYP/6-31G(d) Hydrogen-Bond Strengths (kJ mol-1) between the Watson-Crick (anti) or Hoogsteen (syn) Face of Natural 2′-Deoxyguanosine and the Four Natural DNA Nucleoside Models cytosine
a
EInt ETotb EDef select previously published values
thymine
adenine
guanine
WC
Hoogsteen
WC
Hoogsteen
WC
Hoogsteen
WC
Hoogsteen
-113.5 -100.3 (-93.8) 13.1 105.9,c 120.7,d 107.7,e 112.0,f 116.7,g 99.7h
-37.8 -36.0 (-32.3) 1.8 39.8e
-61.5 -53.1 (-49.1) 8.5 63.3,e 66.6,g 58.2h
-26.4 -24.1 (-21.1) 2.3
-61.8 -53.5 (-47.6) 8.3 66.6,g 54.9h
-28.8 -27.7 (-24.6) 1.1
-48.6 -46.6 (-42.1) 2.0 50.2i
-53.2 -51.2 (-46.1) 2.0
a
The hydrogen-bond strength calculated as the energy difference between the complex and the monomers in the complex geometry (excluding a zero-point vibrational energy correction). b The hydrogen-bond strength calculated as the energy difference between the complex and the individually optimized monomers (excluding a zero-point vibrational energy correction). Binding energies including (0.9806) scaled zero-point vibrational energy corrections are provided in parentheses. c See ref 67. d See ref 68. e See ref 69. f See ref 74. g See ref 83. h See ref 92. i See ref 84.
mismatched pairs (pairs not involving cytosine) with (anti) thymine, (syn) adenine, or (syn) guanine contain only two hydrogen bonds, which are similar to previously reported structures,69,72,76,83,84,92 and the corresponding interaction strengths range from -49 to -62 kJ mol-1 (Table 1). The structural features of the natural dG-C base pair (R(C1′ · · · C1′) ) 10.7 Å and ∠(N9C1′C1′) ) 52.9°) are similar to those expected for DNA, whereas those for the three mismatched pairs show some deviations (Figure 2a). Additionally, propeller-twist and buckle distortions result in a highly nonplanar dG-G base pair. Therefore, despite the magnitude of the calculated binding strengths of (fully optimized) isolated mismatches, their low stability relative to the cytosine pair and their less suitable structure support the observed high specificity of the Watson-Crick face of 2′-deoxyguanosine for hydrogen bonding with cytosine. This discussion shows the importance of considering the interplay between hydrogen-bonding strength and structure in order to predict the binding preference of nucleobases. When 2′-deoxyguanosine rotates about the glycosidic bond to adopt the syn orientation, the Hoogsteen-bonding face becomes available for base-pair formation (Figure 1a, right), which has two hydrogen-bond acceptors (O6 and N7). Similar to previously published results,69 the dG-C complex is significantly destabilized (by 75 kJ mol-1, Table 1) upon hydrogen bonding with the Hoogsteen face of 2′-deoxyguanosine compared to the Watson-Crick face, and is a poor structural match for the natural DNA helix (Figure 2b). This explains why only the Watson-Crick (not the Hoogsteen) dG-C pair is normally observed in DNA and why the anti orientation is preferred. Interestingly, the strongest (EInt) interaction involving the Hoogsteen face of 2′-deoxyguanosine occurs with (anti) guanine (-53 kJ mol-1), followed by cytosine (-38 kJ mol-1), (anti) adenine (-29 kJ mol-1), and thymine (-26 kJ mol-1). Although all of these base pairs deviate from the optimal structure of pairs found in DNA, the structure of the dG-G complex is the closest to B-DNA, which supports previous findings that this Hoogsteen pair is frequently misincorporated.43 The structure of the 2′-deoxyguanosine monomer does not significantly change upon binding with the natural bases. Therefore, when the monomer deformation energy is taken into account by calculating the (ETot) hydrogen-bond strengths as the energy difference between the complexes and the individually optimized monomers, all Watson-Crick and Hoogsteen binding strengths decrease by less than 13 kJ mol-1 and the conclusions drawn from our data hold. Since structural modifications to the natural base may alter the specificity of 2′-deoxyguanosine to hydrogen bond with cytosine, the following section examines the hydrogen-bonding preferences of the
ortho phenoxyl adduct, and discusses their potential influence on the conformational stability of this form of damaged DNA. 2. ortho Adduct. (i) Nucleobase Model. We begin our discussion of the hydrogen-bonded complexes of the ortho phenoxyl adduct by considering the nucleobase model (in the absence of the deoxyribose moiety). However, before the base pairs are analyzed, the structure of the isolated nucleobase adduct must first be discussed, which was carefully studied in our previous work.25 The lowest-energy structure of the nucleobase was determined to be planar with respect to the θ dihedral angle (θ ) 180°, Figure 1), which describes the relative orientation of the natural guanine base and the bulky phenoxyl C8substituent. This planarity results in a strong intramolecular hydrogen bond between the hydroxyl group of the phenoxyl ring and N7 of the guanine nucleobase (Figure 1b). Although other higher-energy conformations of the adduct exist that are twisted with respect to θ,25 we used the lowest-energy (planar) conformation to generate initial guesses for the hydrogen-bonded complexes with the natural nucleobases. It is anticipated that the planar nucleobase is especially desirable in the DNA helix where it would lead to more favorable stacking. Nevertheless, as discussed below, the adduct does not always adopt a planar geometry in fully optimized hydrogen-bonded complexes. In all Watson-Crick pairs involving the (anti) ortho adduct (Figure 3a), the damaged base remains planar (θ ) 180°). Since the Watson-Crick bonding faces of the ortho adduct and natural guanine are identical, the adduct nucleobase forms hydrogen bonds to cytosine with a strength (EInt ) -118 kJ mol-1, Table 2) and structure (Figure 3a) similar to that of 2′-deoxyguanosine (-113 kJ mol-1 (Table 1) and Figure 2a). Although the adduct base pair involving (anti) thymine is planar (Figure 3a, top), the mismatched pairs involving the (syn) purines exhibit propeller-twist and buckle distortions (Figure 3a, bottom), as discussed for the corresponding natural pairs. The (EInt) binding strengths of the three mismatched pairs range from -55 to -65 kJ mol-1 (Table 2). Despite being significant in magnitude, these binding arrangements are only half as stable as the adduct-cytosine base pair. Furthermore, there is little difference between the strength of pairs involving the adduct (Table 2) or the natural base (Table 1). These results for the nucleobase model indicate that the addition of the bulky group to C8 does not affect the Watson-Crick hydrogen-bonding pattern of guanine and therefore the Watson-Crick face of the adduct will be as selective in base pairing as natural 2′-deoxyguanosine. Similar to natural guanine, the Hoogsteen face of the (syn) ortho phenoxyl adduct has two hydrogen-bond acceptors (O6 and N7, Figure 1b). However, the hydroxyl group on the phenyl ring provides a third hydrogen-bonding site, which can act as a
Stability of C8-Phenoxyl-2′-deoxyguanosine Adducts
Figure 3. B3LYP/6-31G(d) structures (distances in Å, angles in deg) for complexes between the (a) Watson-Crick (anti) and (b) Hoogsteen (syn) faces of the ortho C8-phenoxyl-guanine (nucleobase) adduct and each of the four natural DNA nucleobases.
hydrogen-bond donor or acceptor depending on the hydroxyl orientation. The direct participation of the bulky substituent in hydrogen bonding results in a Hoogsteen pair with cytosine (Figure 3b) that is stabilized (by 64 kJ mol-1) relative to the corresponding pair with natural 2′-deoxyguanosine. Furthermore, the Hoogsteen face of the adduct paired with cytosine is only slightly less stable (by 16 kJ mol-1) than the corresponding Watson-Crick pair (Table 2). In comparison, the difference in the strengths of these two cytosine hydrogen-bonding patterns with natural 2′-deoxyguanosine is 75 kJ mol-1. The comparable stabilities of the adduct-cytosine Hoogsteen and Watson-Crick interactions indicates that hydrogen bonding with cytosine in DNA will not significantly destabilize the syn orientation of the adduct relative to the anti orientation. Therefore, unlike natural 2′-deoxyguanosine, the presence of the bulky C8 group may afford two favorable conformations of the adduct when paired with cytosine. Nevertheless, the quality of the adductcytosine Hoogsteen pair is dependent on the structural fit within a DNA helix, which is difficult to evaluate using the nucleobase model discussed in this section. This emphasizes the importance of using a larger model for determining base-pairing preferences. When mismatches are considered, the next strongest Hoogsteen bonding mode of the ortho phenoxyl adduct occurs with guanine (EInt ) -56 kJ mol-1, Figure 3b), which is the strongest pair with syn natural 2′-deoxyguanosine. However, thymine and adenine both form very weak Hoogsteen pairs with the ortho adduct (-16 and -20 kJ mol-1, respectively). These mismatches are both slightly destabilized relative to the corresponding natural 2′-deoxyguanosine pairs, and are much weaker than the natural dG-C (-113.5 kJ mol-1) and dA-T (-54.1 kJ mol-1)93 contacts. Furthermore, as found for natural 2′-deoxyguanosine,
J. Phys. Chem. B, Vol. 114, No. 40, 2010 12999 the propeller-twist and buckle distortions are significant when the adduct binds to thymine or guanine, which suggests that neither pair will properly fit into a helix. This suggests that, unlike the cytosine Hoogsteen pair, the presence of the three possible mismatches will likely destabilize DNA helices. Thus, if the ortho adduct appears in the syn orientation, then a strong Hoogsteen complex will only be formed with cytosine. It is particularly interesting that the ortho adduct binds strongly to cytosine regardless of whether the anti or syn conformation is adopted. This may indicate that the adduct can exist in either conformation in a DNA helix without causing destabilization or mutations. However, closer examination of the geometry of the adduct monomer reveals larger structural changes when bound to cytosine using the Hoogsteen face compared with the Watson-Crick face. Specifically, the monomer in the Watson-Crick pair resembles the global minimum of the (isolated) ortho nucleobase adduct, which is planar and strongly stabilized by the presence of an O-H · · · N7 intramolecular hydrogen bond.25 However, in the Hoogsteen base pair, the hydroxyl group in the adduct rotates and causes a twisted conformation about θ due to a repulsive H-O · · · N7 interaction, which corresponds to a significantly higher-energy conformation of the nucleobase (by 52 kJ mol-1 relative to the planar global minimum).25 Since the adduct geometry is significantly affected by base-pair formation, the deformation energy must be examined in order to accurately compare the relative hydrogen-bond strengths. As found for natural 2′-deoxyguanosine (Table 1), the adduct geometry is not greatly affected by Watson-Crick base-pair formation. Therefore, the hydrogen-bond strengths decrease by only 2-13 kJ mol-1 when the deformation energy is included in the interaction energy (Table 2). Specifically, the (ETot) binding strength of the adduct-cytosine pair decreases to -105 kJ mol-1, while those of the mismatched pairs with thymine, adenine, and guanine fall within a range of -53 to -57 kJ mol-1 (Table 2). This confirms that cytosine leads to the strongest Watson-Crick pair, and our previous conclusions hold regardless of whether the deformation energy is included in the calculated Watson-Crick hydrogen-bond strengths. The effect of accounting for the monomer deformation energy in the total interaction energy is much larger for the Hoogsteen pairs of the ortho adduct than for the corresponding Watson-Crick pairs or the natural Hoogsteen pairs. For example, the total (ETot) binding strength of the cytosine pair is only -27 kJ mol-1, which represents a decrease in stability of 74 kJ mol-1 due to rotation of the hydroxyl group in the adduct upon binding (Table 2). The adduct-guanine pair leads to a similar conformational change and a significant decrease in the binding energy. In fact, when the deformation energy is included, the adduct-guanine interaction is (1.5 kJ mol-1) destabilizing even though the natural 2′-deoxyguanosine pair remains (-51 kJ mol-1) stable. The thymine and adenine interactions with the adduct are only slightly affected by the deformation energy and remain weakly favorable (-15 and -18 kJ mol-1, respectively). Therefore, even when the deformation of the adduct monomer is taken into account, cytosine interactions with the Hoogsteen face are the strongest among all bases (Table 2). Nevertheless, the (ETot) binding strength of this pair is much weaker (by 27 kJ mol-1) than a typical 2′-deoxyadenosine-thymine interaction (-54.1 kJ mol-1).93 It is also interesting to note that the deformation energy for the natural 2′-deoxyguanosine-cytosine pair is less than 2 kJ mol-1, and therefore the total cytosine interaction strength (including deformation energy) for the adduct (-27 kJ mol-1) is now weaker than that for the natural nucleoside
13000
J. Phys. Chem. B, Vol. 114, No. 40, 2010
Millen et al.
TABLE 2: B3LYP/6-311+G(2df,p)//B3LYP/6-31G(d) Hydrogen-Bond Strengths (kJ mol-1) between the Watson-Crick (anti) or Hoogsteen (syn) Face of the ortho or para C8-Phenoxyl-guanine (Nucleobase) Adduct and the Four Natural DNA Nucleobases cytosine a
EInt
ETotc
ortho para ortho para
EDef
ortho para
thymine
adenine
guanine
WC
Hoogsteen
WC
Hoogsteen
WC
Hoogsteen
WC
Hoogsteen
-118.1 -114.9 -105.2 (-98.9) -102.0 (-95.8) 12.9 12.9
-101.8 -57.4 -27.2 (-24.0) -53.2 (-48.0) 74.7 4.3
-64.8 -64.0 -56.9 (-52.8) -56.0 (-51.9) 7.8 8.0
-16.3 -b -14.9 (-12.5) -b
-61.1 -60.7 -53.4 (-47.4) -53.1 (-47.1) 7.7 7.6
-20.4 -35.3 -17.6 (-15.0) -32.7 (-28.4) 2.9 2.7
-54.9 -49.6 -52.9 (-48.5) -47.7 (-43.4) 2.0 1.9
-56.1 -49.9 1.5 (4.7) -47.4 (-42.8) 57.6 2.5
1.4 -b
a The hydrogen-bond strength calculated as the energy difference between the complex and the monomers in the complex geometry (excluding a zero-point vibrational energy correction). b No viable thymine pair was identified (see ref 95). c The hydrogen-bond strength calculated as the energy difference between the complex and the individually optimized monomers (excluding a zero-point vibrational energy correction). Binding energies including (0.9806) scaled zero-point vibrational energy corrections are provided in parentheses.
(-36 kJ mol-1). Thus, it is anticipated that any pair formed with the Hoogsteen face of the adduct will destabilize DNA relative to the corresponding natural sequence. In summary, when the ortho phenoxyl adduct monomer deformation energy is included in the calculated interaction strengths, the Hoogsteen pairs are greatly destabilized relative to the Watson-Crick pairs. Since the anti/syn conformational energy difference was previously calculated to be small in nucleotide systems,30 it is anticipated that the strong cytosine Watson-Crick interaction compared to the Hoogsteen interaction will lead to a preference for the anti orientation of the ortho phenoxyl adduct in DNA. In addition, even when mismatches are considered, the anti orientation of the ortho adduct will form stronger base pairs than the syn conformation. This suggests that the potential for hydrogen bonding to induce conformational changes in this adduct relative to the corresponding natural base, as well as the potential for this particular form of damage to mispair, is small. Therefore, the ortho adduct may exert its biological effects in other ways. (ii) Nucleoside Model. The preceding discussion indicates that deformation of the adduct monomer upon base-pair formation has a large effect on the calculated hydrogen-bond strength. However, our previous work shows that the monomer geometry also changes in the presence of the DNA sugar moiety.25 Specifically, inclusion of deoxyribose induces a twist about the θ dihedral angle by 21° in the lowest energy anti conformation of the isolated ortho nucleoside adduct and 25° in the lowest energy syn structure.25 Since it is unknown how the twist will affect hydrogen-bond formation or vice versa, we consider the effects of using a nucleoside model on the calculated hydrogen-bonded structures and strengths. The structures of the Watson-Crick (anti orientation) pairs involving the adduct nucleoside are very similar to the corresponding natural complexes, where the complex with cytosine has a geometry most suitable for a typical B-DNA helix (Figure 4a). More importantly, the ortho adduct nucleoside remains twisted about θ by ∼21° in all Watson-Crick hydrogen-bonded complexes compared to θ ) 180° (planar) when the nucleobase model is considered (Figure 3a). Nevertheless, the (EInt) interaction energies of the two models deviate by less than 2 kJ mol-1 (Tables 2 and 3), which indicates that the (EInt) Watson-Crick hydrogen-bond strengths are not affected by the degree of twist. Therefore, cytosine remains the strongest Watson-Crick bonding partner upon model extension (Table 3). Furthermore, the effect of including the monomer deformation energy is similar for the nucleoside and nucleobase models, as well as for all base pairs. Specifically, the total (ETot) Watson-Crick hydrogen-
Figure 4. B3LYP/6-31G(d) structures (distances in Å, angles in deg) for complexes between the (a) Watson-Crick (anti) and (b) Hoogsteen (syn) faces of the ortho C8-phenoxyl-2′-deoxyguanosine (nucleoside) adduct and each of the four natural DNA nucleoside models.
bond strength for each nucleoside base pair is smaller by only 1-12 kJ mol-1 (Table 2), where the most significant difference occurs for the cytosine complex. Therefore, as found for the nucleobase model, the hydrogen-bond strengths calculated using the adduct nucleoside follow the same trend as those for the natural nucleoside and support preferential pairing with cytosine. The (EInt) hydrogen-bond strengths of pairs involving the Hoogsteen face (syn orientation) of the ortho nucleoside (Table
Stability of C8-Phenoxyl-2′-deoxyguanosine Adducts
J. Phys. Chem. B, Vol. 114, No. 40, 2010 13001
TABLE 3: B3LYP/6-311+G(2df,p)//B3LYP/6-31G(d) Hydrogen-Bond Strengths (kJ mol-1) between the Watson-Crick (anti) or Hoogsteen (syn) Face of the ortho or para C8-Phenoxyl-2′-deoxyguanosine (Nucleoside) Adduct and the Four Natural DNA Nucleoside Models cytosine a
EInt
ETotc
ortho para ortho para
EDef
ortho para
thymine
adenine
guanine
WC
Hoogsteen
WC
Hoogsteen
WC
Hoogsteen
WC
Hoogsteen
-117.3 -113.9 -105.4 (-99.2) -100.6 (-94.7) 11.9 13.3
-98.9 -54.3 -38.6 (-35.9) -49.8 (-44.2) 60.2 4.5
-62.7 -61.8 -55.8 (-52.0) -53.7 (-50.0) 6.9 8.1
-16.8 -b -15.3 (-12.8) -b
-62.6 -61.9 -55.5 (-49.7) -53.9 (-48.3) 7.0 8.0
-21.1 -34.2 -16.6 (-13.8) -31.5 (-26.6) 4.5 2.7
-55.5 -50.1 -54.4 (-50.0) -48.1 (-43.7) 1.1 2.0
-57.5 -53.7 -12.3 (-10.1) -51.2 (-46.4) 45.2 2.5
1.5 -b
a The hydrogen-bond strength calculated as the energy difference between the complex and the monomers in the complex geometry (excluding a zero-point vibrational energy correction). b No viable thymine pair was identified (see ref 95). c The hydrogen-bond strength calculated as the energy difference between the complex and the individually optimized monomers (excluding a zero-point vibrational energy correction). Binding energies including (0.9806) scaled zero-point vibrational energy corrections are provided in parentheses.
3) are smaller by up to only 3 kJ mol-1 than the adduct nucleobase (Table 2). Furthermore, the geometries of the complexes determined using the nucleoside (Figure 4b) and nucleobase (Figure 3b) models are similar with the exception of the twist about θ. Interestingly, the degree of twist depends on the model and the base pair considered, which implies that inclusion of the deformation energy will have a varying effect on the calculated binding strengths of some nucleoside pairs compared with the nucleobase model. For example, in the cytosine complex, the nucleoside adduct is twisted about θ by 40°, which leads to a decrease in the hydrogen-bond strength of 60 kJ mol-1 (Table 3) upon inclusion of monomer deformation energy, while the nucleobase adduct is twisted by 26°, which results in a 74 kJ mol-1 reduction in binding (Table 2). This difference leads to a 11 kJ mol-1 stronger (ETot) total interaction energy with cytosine for the nucleoside model compared with the nucleobase model. Although the thymine and adenine complexes are not significantly affected by the model or deformation energy, inclusion of the deformation energy leads to a 14 kJ mol-1 stronger (ETot) binding strength for the adduct-guanine pair when the larger nucleoside model is used. Comparison of binding strengths that neglect (EInt) and include (ETot) the monomer deformation energy indicate that changes in the monomer geometry upon binding have a significant effect on the overall base-pair hydrogen-bond strength. This effect is in part due to the large conformational (twist) flexibility in the adduct, which is dependent upon the presence of the deoxyribose sugar moiety. Unfortunately, the exact structure of the adduct in single- versus double-stranded DNA, which would afford the deformation energy upon binding in DNA environments, is currently unknown. Although expanding the computational model to yield structures even more similar to those found in DNA will further improve the accuracy of the calculated deformation energy, it is realistic to expect some nucleoside deformation to occur upon formation of double helices. Since the sugar moiety has a large effect on θ, we believe that the nucleoside model provides a better estimate of the deformation energy than the nucleobase model. Therefore, despite the popularity of using small, nucleobase models to study hydrogenbonding patterns,60,70,74-77,85,90,92 we conclude that future work should consider adduct models at least as large as the nucleoside in order to calculate accurate interaction energies in DNA. This is especially true for adducts that may undergo large structural changes with attachment of the deoxyribose sugar moiety and/ or upon inclusion into DNA helices.
The Hoogsteen mismatches involving the ortho phenoxyl adduct nucleoside are significantly destabilized relative to the corresponding pairs with the natural nucleoside upon inclusion of the monomer deformation energy. In fact, the (ETot) hydrogenbond strengths with adenine, guanine, or thymine are diminished to the point where these pairs are not stable enough to be feasible in a DNA helix (Table 3). On the other hand, the adduct-cytosine complex is actually slightly (2.6 kJ mol-1) more stable than the natural dG-C pair. Furthermore, the R(C1′ · · · C1′) distance and the ∠(N9C1′C1′) angle in the adduct complex are closer to the values expected for DNA than the natural nucleoside pair. Therefore, despite the similarity in stability of the Hoogsteen cytosine pair involving the ortho adduct and 2′-deoxyguanosine, the shape of the adduct complex is typical of a natural Watson-Crick pair, and the syn orientation of the adduct may consequently be less destabilizing than that of the natural nucleoside when paired with cytosine in DNA helices. This example illustrates that the presence of the bulky group can improve the Hoogsteen binding and reemphasizes the importance of considering both the binding strength and structure when evaluating base pairs, as well as the use of larger models that afford more detailed structural information. In summary, as found for the nucleobase model, deformation of the ortho phenoxyl adduct nucleoside upon base-pair formation has a large effect on the calculated hydrogen-bond strengths, where all base pairs are predicted to be less stable when the deformation energy is included. In addition to the valuable structural information obtained from the larger model, the significant differences in the deformation energy between the two models provides further support for using at least a nucleoside model in future studies of other flexible, bulky adducts. Given that the Watson-Crick interaction is significantly stronger (by 67 kJ mol-1), the anti conformation of the damaged base may be stabilized in the helix. Nevertheless, the syn orientation also forms a reasonably strong (-38.6 kJ mol-1), and structurally viable, base pair with cytosine, and may still afford a stable double helix should other factors (stacking interactions, major groove contacts) induce the syn orientation. Most importantly, cytosine is the base favored by the ortho nucleoside adduct regardless of the orientation about the glycosidic bond. Therefore, hydrogen-bonding preferences suggest that misincorporation is not expected to occur for the ortho phenoxyl adduct. 3. para Adduct. Since we previously concluded that the nucleoside model is required to accurately calculate the binding strengths and provide detailed structural information of the ortho
13002
J. Phys. Chem. B, Vol. 114, No. 40, 2010
adduct pairs, only the results obtained with the nucleoside model for the para phenoxyl adduct will be discussed in this section. In addition, only the total (ETot) binding strengths for the para adduct complexes that include the monomer deformation energy will be discussed, since these values for the ortho adduct were concluded to be the most relevant to DNA environments. We note that the interaction energies without the deformation energy included for the nucleoside model are also available in Table 3. Additionally, the binding strengths and geometries calculated with the nucleobase model can be found in Table 2 and Figure S1 (Supporting Information), respectively, which show that the smaller model yields binding strengths and geometries similar to the nucleoside model for the para adduct.94 The para phenoxyl adduct lacks the stabilizing O-H · · · N7 intramolecular hydrogen bond contained in the ortho adduct (Figure 1), which leads to a larger twist in the isolated nucleoside with respect to θ in both the anti and syn orientations (θ ) 31 or 34° for the para compared to 21 or 25° for the ortho adduct anti or syn conformations, respectively).30 As discussed for the ortho adduct, deviations in the twist angle can change the geometries of the hydrogen-bonding pairs. It is not intuitive how differences in the twist in the ortho and para monomers, as well as the different location of the hydroxyl group, will affect the overall base-pairing preference of the damaged base. The anti orientation of the para phenoxyl adduct is expected to form hydrogen-bonded pairs with the natural nucleosides in an analogous manner to the ortho adduct, since the Watson-Crick face is unaffected by the damage. Indeed, the two adducts form structurally similar hydrogen-bonding interactions with each natural nucleoside model (Figures 4a and 5a) and therefore the calculated binding strengths are within 6 kJ mol-1 (Table 3). The largest deviation occurs between the guanine pairs due to greater propeller-twist and buckle distortions in the para complex. All structures and binding strengths are also close to those for the corresponding natural 2′-deoxyguanosine pairs. Most importantly, as found for natural 2′-deoxyguanosine and the ortho adduct, the para adduct interacts most strongly with cytosine, which also forms the pair with the most appropriate structure for a DNA helix. Unlike the ortho adduct, the Hoogsteen-bonding face of the (syn) para phenoxyl adduct (Figure 1c) has no additional hydroxyl hydrogen-bonding sites compared with natural guanine. Instead, the bulky phenoxyl group forms additional weak C-H · · · X interactions upon binding to the natural bases. These interactions actually preclude the formation of a viable thymine pair.95 Furthermore, the para adduct has slightly larger Hoogsteen binding strengths with the other natural bases than 2′deoxyguanosine (by up to 13.8 kJ mol-1). The largest energetic difference occurs for the cytosine pair, which also has a significant (R(C1′ · · · C1′), ∠(N9C1′C1′)) structural improvement (10.5 Å, 45°) over the (syn) natural dG complex (11.7 Å, 30°). Although the adduct-adenine pair is distorted, structural improvement is also seen in this complex compared to the 2′deoxyguanosine mismatch. Therefore, the presence of the bulky group improves the quality of these Hoogsteen complexes relative to the corresponding pair composed of natural bases. As found for natural 2′-deoxyguanosine, the guanine Hoogsteen complex has the largest binding strength despite significant propeller-twist and buckle distortions (Figure 5b). However, for the para phenoxyl adduct, the cytosine complex is similar in stability and is less distorted than the guanine complex. Since calculations on other non-natural base pairs indicate that calculated binding strengths may not be significantly affected even if environmental constraints enforce a less distorted
Millen et al.
Figure 5. B3LYP/6-31G(d) structures (distances in Å, angles in deg) for complexes between the (a) Watson-Crick (anti) and (b) Hoogsteen (syn) faces of the para C8-phenoxyl-2′-deoxyguanosine (nucleoside) adduct and each of the four natural DNA nucleoside models.
complex,96 our results indicate that both cytosine and guanine may form viable Hoogsteen pairs of similar stability (Table 3). Therefore, if the para phenoxyl adduct adopts the syn orientation, favorable mispairing with guanine may result in possible mutagenic effects for this adduct. Interestingly, all Hoogsteen pairs are more stable than the corresponding ortho phenoxyl adduct complexes (by 11-39 kJ mol-1). Thus, while the Hoogsteen complexes of the ortho adduct are quite unstable relative to the Watson-Crick pairs, there is a smaller difference in stability between the para Hoogsteen and Watson-Crick complexes (Table 3). These results clearly suggest a strong preference for all nucleobases to form Watson-Crick pairs with the ortho adduct but cannot undisputedly predict the preference for the para adduct. Indeed, if other factors (stacking, major groove contacts) cause the phenoxyl adducts to adopt the syn orientation in the DNA helix, the para adduct may be less destabilizing than the ortho adduct and syn may even become the preferred orientation. In summary, among all natural bases, cytosine forms the most structurally feasible pairs with both the Watson-Crick and Hoogsteen faces of the para phenoxyl adduct, where the Watson-Crick complex is more stable. However, both cytosine and guanine form relatively stable Hoogsteen pairs that are comparable in strength to the natural adenine-thymine complex.
Stability of C8-Phenoxyl-2′-deoxyguanosine Adducts Furthermore, although Watson-Crick binding is strongly favored for natural 2′-deoxyguanosine and the ortho adduct, the difference between the Watson-Crick and Hoogsteen binding faces is significantly smaller for the para adduct. Indeed, if the phenoxyl adducts adopt the syn orientation in the DNA helix, the para adduct may be less destabilizing than the natural nucleoside or the ortho adduct. Therefore, the anti/syn preference for the para adduct is less clear than that for the ortho adduct, and the para adduct may readily form base pairs in either orientation with more than one natural base. We propose that simulations on damaged DNA strands are needed to unambiguously determine the conformational and base-pairing preference of the para adduct within the constraints of the DNA duplex environment, which will better reveal how this adduct could lead to mutations. Conclusion The hydrogen-bonding preferences of the Watson-Crick (anti) and Hoogsteen (syn) faces of the ortho and para C8phenoxyl-2′-deoxyguanosine adducts for the four natural bases were calculated to determine the possible effect of hydrogenbond formation on the anti/syn preference of the damaged bases, as well as predict the potential for misincorporation events upon replication. The magnitudes of the binding strengths, as well as the structural features of the complexes (R(C1′ · · · C1′) distance, ∠(N9C1′C1′) angle, propeller-twist and buckle distortions), were evaluated and compared to those expected for natural DNA helices. Comparison of data from nucleoside and nucleobase models of the ortho phenoxyl adduct suggests that the nucleobase model overestimates the magnitude of the deformation energy. Therefore, even though both models have the same base-pairing trend when deformation energy is included, nucleoside models are the smallest computational models that should be used for accurate prediction of the binding strengths of flexible, bulky adducts that may undergo significant geometric changes upon complex formation. Using our most accurate models to date, we conclude that the ortho phenoxyl adduct prefers to bind to cytosine regardless of the hydrogen-bonding face considered, which suggests this form of damage will have a low potential to result in misincorporation. Since previous results indicate the anti/syn energy difference is small30 and the present work indicates the Watson-Crick face forms significantly more stable complexes than the Hoogsteen face, the anti conformation of the adduct will be favored by hydrogen-bonding interactions in DNA helices. However, the Hoogsteen face also forms a reasonably strong and structurally feasible pair with cytosine. Therefore, this type of damage is not expected to lead to misincorporation even if the syn orientation is adopted in DNA helices due to other environmental factors. While the ortho adduct and natural nucleosides form much stronger complexes with the Watson-Crick than the Hoogsteen face, the difference in stability is less significant for the para adduct. This suggests that the para adduct may readily adopt multiple conformations in DNA helices. Furthermore, although the Watson-Crick face forms the most stable complex with cytosine, the Hoogsteen face forms hydrogen bonds to both cytosine and guanine that are comparable in stability to a Watson-Crick adenine-thymine pair. Therefore, the syn orientation of the para adduct may be adopted in DNA helices if other factors favor this orientation, and this damaged base may bind to multiple partners in this conformation. This suggests a mechanism by which this type of DNA damage may lead to mutations. Nevertheless, the isolated para adduct-guanine
J. Phys. Chem. B, Vol. 114, No. 40, 2010 13003 complex is significantly distorted and the effects of other environmental considerations in the DNA strand (such as stacking, major groove contacts, etc.) on the binding strength must be investigated to conclusively determine the conformational and binding preference of this type of damaged base. Acknowledgment. We thank the Natural Sciences and Engineering Research Council (NSERC), the Canada Foundation for Innovation (CFI), and the Canada Research Chair (CRC) program for financial support. A.L.M. thanks NSERC, Alberta Ingenuity (now part of Alberta Innovates - Technology Futures), the Alberta government, and the University of Lethbridge for graduate student scholarships. Supporting Information Available: Figure showing the B3LYP/6-31G(d) structures for complexes between the WatsonCrick (anti) and Hoogsteen (syn) faces of the para C8-phenoxylguanine (nucleobase) adduct and each of the four natural DNA nucleobases. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Nakabeppu, Y.; Tsuchimoto, D.; Yamaguchi, H.; Sakumi, K. J. Neurosci. Res. 2007, 85, 919–934. (2) Drabløs, F.; Feyzi, E.; Aas, P. A.; Vaagbø, C. B.; Kavli, B.; Bratlie, M. S.; Pen˜a-Diaz, J.; Otterlei, M.; Slupphaug, G.; Krokan, H. E. DNA Repair 2004, 3, 1389–1407. (3) Labet, V.; Grand, A.; Morell, C.; Cadet, J.; Eriksson, L. A. Theor. Chem. Acc. 2008, 120, 429–435. (4) Kow, Y. W. Free Radical Biol. Med. 2002, 33, 886–893. (5) Boysen, G.; Pachkowski, B. F.; Nakamura, J.; Swenberg, J. A. Mutat. Res., Genet. Toxicol. EnViron. Mutagen. 2009, 678, 76–94. (6) Gannett, P. M.; Lawson, T.; Miller, M.; Thakkar, D. D.; Lord, J. W.; Yau, W. M.; Toth, B. Chem.-Biol. Interact. 1996, 101, 149–164. (7) Parks, J. M.; Ford, G. P.; Cramer, C. J. J. Org. Chem. 2001, 66, 8997–9004. (8) Dai, J.; Wright, M. W.; Manderville, R. A. Chem. Res. Toxicol. 2003, 16, 817–821. (9) Dai, J.; Wright, M. W.; Manderville, R. A. J. Am. Chem. Soc. 2003, 125, 3716–3717. (10) Heavner, S.; Gannett, P. M. J. Biomol. Struct. Dyn. 2005, 23, 203– 219. (11) Stover, J. S.; Ciobanu, M.; Cliffel, D. E.; Rizzo, C. J. J. Am. Chem. Soc. 2007, 129, 2074–2081. (12) Gannett, P. M.; Heavner, S.; Daft, J. R.; Shaughnessy, K. H.; Epperson, J. D.; Greenbaun, N. L. Chem. Res. Toxicol. 2003, 16, 1385– 1394. (13) Yang, Z. Z.; Qi, S. F.; Zhao, D. X.; Gong, L. D. J. Phys. Chem. B 2009, 113, 254–259. (14) Guengerich, F. P. Drug Metab. ReV. 2002, 34, 607–623. (15) Kikugawa, K.; Kato, T.; Kojima, K. Mutat. Res. 1992, 268, 65– 75. (16) Dai, Q.; Xu, D. W.; Lim, K.; Harvey, R. G. J. Org. Chem. 2007, 72, 4856–4863. (17) Faucet, V.; Pfohl-Leszkowicz, A.; Dai, J.; Castegnaro, M.; Manderville, R. A. Chem. Res. Toxicol. 2004, 17, 1289–1296. (18) Dai, J.; Wright, M. W.; Manderville, R. A. J. Am. Chem. Soc. 2003, 125, 3716–3717. (19) Manderville, R. A. Chem. Res. Toxicol. 2005, 18, 1091–1097. (20) Pfohl-Leszkowicz, A.; Manderville, R. A. Mol. Nutr. Food Res. 2007, 51, 61–99. (21) Tozlovanu, M.; Faucet-Marquis, V.; Pfohl-Leszkowicz, A.; Manderville, R. A. Chem. Res. Toxicol. 2006, 19, 1241–1247. (22) Mantle, P. G.; Faucet-Marquis, V.; Manderville, R. A.; Squillaci, B.; Pfohl-Leszkowicz, A. Chem. Res. Toxicol. 2010, 23, 89–98. (23) Sun, K. M.; McLaughlin, C. K.; Lantero, D. R.; Manderville, R. A. J. Am. Chem. Soc. 2007, 129, 1894–1895. (24) McLaughlin, C. K.; Lantero, D. R.; Manderville, R. A. J. Phys. Chem. A 2006, 110, 6224–6230. (25) Millen, A. L.; McLaughlin, C. K.; Sun, K. M.; Manderville, R. A.; Wetmore, S. D. J. Phys. Chem. A 2008, 112, 3742–3753. (26) Dai, J.; Sloat, A. L.; Wright, M. W.; Manderville, R. A. Chem. Res. Toxicol. 2005, 18, 771–779. (27) Weishar, J. L.; McLaughlin, C. K.; Baker, M.; Gabryelski, W.; Manderville, R. A. Org. Lett. 2008, 10, 1839–1842. (28) Manderville, R. A. Can. J. Chem. 2005, 83, 1261–1267.
13004
J. Phys. Chem. B, Vol. 114, No. 40, 2010
(29) Schlitt, K. M.; Sun, K. W. M.; Paugh, R. J.; Millen, A. L.; NavarroWhyte, L.; Wetmore, S. D.; Manderville, R. A. J. Org. Chem. 2009, 74, 5793–5802. (30) Millen, A. L.; Manderville, R. A.; Wetmore, S. D. J. Phys. Chem. B 2010, 114, 4373–4382. (31) Mao, B.; Vyas, R. R.; Hingerty, B. E.; Broyde, S.; Basu, A. K.; Patel, D. J. Biochemistry 1996, 35, 12659–12670. (32) Eason, R. G.; Burkhardt, D. M.; Phillips, S. J.; Smith, D. P.; David, S. S. Nucleic Acids Res. 1996, 24, 890–897. (33) Burnouf, D. Y.; Wagner, J. E. J. Mol. Biol. 2009, 386, 951–961. (34) Hilario, P.; Yan, S. X.; Hingerty, B. E.; Broyde, S.; Basu, A. K. J. Biol. Chem. 2002, 277, 45068–45074. (35) Jin, B.; Lee, H. M.; Kim, S. K. J. Biomol. Struct. Dyn. 2010, 27, 457–464. (36) Watt, D. L.; Utzat, C. D.; Hilario, P.; Basu, A. K. Chem. Res. Toxicol. 2007, 20, 1658–1664. (37) Brown, K.; Hingerty, B. E.; Guenther, E. A.; Krishnan, V. V.; Broyde, S.; Turteltaub, K. W.; Cosman, M. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 8507–8512. (38) Wang, F.; DeMuro, N. E.; Elmquist, C. E.; Stover, J. S.; Rizzo, C. J.; Stone, M. P. J. Am. Chem. Soc. 2006, 128, 10085–10095. (39) Gubala, V.; Betancourt, J. E.; Rivera, J. M. Org. Lett. 2004, 6, 4735–4738. (40) Sessler, J. L.; Jayawickramarajah, J.; Sherman, C. L.; Brodbelt, J. S. J. Am. Chem. Soc. 2004, 126, 11460–11461. (41) DeCarlo, L.; Gowda, A. S. P.; Suo, Z. C.; Spratt, T. E. Biochemistry 2008, 47, 8157–8164. (42) Johnson, R. E.; Haracska, L.; Prakash, L.; Prakash, S. Mol. Cell. Biol. 2006, 26, 6435–6441. (43) Benitez, B. A. S.; Arora, K.; Balistreri, L.; Schlick, T. J. Mol. Biol. 2008, 384, 1086–1097. (44) Gannett, P. M.; Sura, T. P. Chem. Res. Toxicol. 1993, 6, 690–700. (45) Shapiro, R.; Ellis, S.; Hingerty, B. E.; Broyde, S. Chem. Res. Toxicol. 1998, 11, 335–341. (46) Patel, D. J.; Mao, B.; Gu, Z. T.; Hingerty, B. E.; Gorin, A.; Basu, A. K.; Broyde, S. Chem. Res. Toxicol. 1998, 11, 391–407. (47) Meneni, S. R.; Shell, S. M.; Gao, L.; Jurecka, P.; Lee, W.; Sponer, J.; Zou, Y.; Chiarelli, M. P.; Cho, B. P. Biochemistry 2007, 46, 11263– 11278. (48) Wang, F.; Elmquist, C. E.; Stover, J. S.; Rizzo, C. J.; Stone, M. P. Biochemistry 2007, 46, 8498–8516. (49) Singh, A. K.; Mishra, P. C. Int. J. Quantum Chem. 2005, 102, 343– 351. (50) Jia, L.; Shafirovich, V.; Shapiro, R.; Geacintov, N. E.; Broyde, S. Biochemistry 2006, 45, 6644–6655. (51) Cho, B. S. P. J. EnViron. Sci. Health, Part C 2004, 22, 57–90. (52) Jia, L.; Shafirovich, V.; Shapiro, R.; Geacintov, N. E.; Broyde, S. Biochemistry 2005, 44, 6043–6051. (53) Bichara, M.; Fuchs, R. P. P. J. Mol. Biol. 1985, 183, 341–351. (54) Gu, Z. T.; Gorin, A.; Hingerty, B. E.; Broyde, S.; Patel, D. J. Biochemistry 1999, 38, 10855–10870. (55) Cho, B. P.; Beland, F. A.; Marques, M. M. Biochemistry 1992, 31, 9587–9602. (56) Hernández, B.; Soliva, R.; Luque, F. J.; Orozco, M. Nucleic Acids Res. 2000, 28, 4873–4883. (57) Hu, X. B.; Li, H. R.; Ding, J. Y.; Han, S. J. Biochemistry 2004, 43, 6361–6369. (58) Volk, D. E.; Thiviyanathan, V.; Somasunderam, A.; Gorenstein, D. G. Org. Biomol. Chem. 2007, 5, 1554–1558. (59) Sˇponer, J.; Sabat, M.; Gorb, L.; Leszczynski, J.; Lippert, B.; Hobza, P. J. Phys. Chem. B 2000, 104, 7535–7544. (60) Sˇponer, J.; Jurecka, P.; Hobza, P. J. Am. Chem. Soc. 2004, 126, 10142–10151. (61) Fonseca Guerra, C.; Bickelhaupt, F. M.; Snijders, J. G.; Baerends, E. J. J. Am. Chem. Soc. 2000, 122, 4117–4128. (62) Giudice, E.; Várnai, P.; Lavery, R. Nucleic Acids Res. 2003, 31, 1434–1443. (63) Boys, S. F.; Bernardi, F. Mol. Phys. 1970, 19, 553–566. (64) Scott, A. P.; Radom, L. J. Phys. Chem. 1996, 100, 16502–16513. (65) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.;
Millen et al. Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revisions C.02 and D.01; Gaussian, Inc.: Wallingford, CT, 2004. (66) Deepa, P.; Kolandaivel, P. J. Biomol. Struct. Dyn. 2008, 25, 733– 746. (67) Gould, I. R.; Kollman, P. A. J. Am. Chem. Soc. 1994, 116, 2493– 2499. (68) Jurecka, P.; Hobza, P. J. Am. Chem. Soc. 2003, 125, 15608–15613. (69) Kabelácˇ, M.; Hobza, P. J. Phys. Chem. B 2001, 105, 5804–5817. (70) Kryachko, E. S.; Sabin, J. R. Int. J. Quantum Chem. 2003, 91, 695–710. (71) Liu, H.; Gauld, J. W. Phys. Chem. Chem. Phys. 2009, 11, 278– 287. (72) Mathews, D. H.; Case, D. A. J. Mol. Biol. 2006, 357, 1683–1693. (73) Mo, Y. R. J. Mol. Model. 2006, 12, 665–672. (74) Mohajeri, A.; Nobandegani, F. F. J. Phys. Chem. A 2008, 112, 281– 295. (75) Noguera, M.; Sodupe, M.; Bertran, J. Theor. Chem. Acc. 2004, 112, 318–326. (76) Padermshoke, A.; Katsumoto, Y.; Masaki, R.; Aida, M. Chem. Phys. Lett. 2008, 457, 232–236. (77) Quinn, J. R.; Zimmerman, S. C.; Del Bene, J. E.; Shavitt, I. J. Am. Chem. Soc. 2007, 129, 934–941. (78) Radisic, D.; Bowen, K. H.; Da¸bkowska, I.; Storoniak, P.; Rak, J.; Gutowski, M. J. Am. Chem. Soc. 2005, 127, 6443–6450. (79) Sharma, P.; Mitra, A.; Sharma, S.; Singh, H.; Bhattacharyya, D. J. Biomol. Struct. Dyn. 2008, 25, 709–732. (80) Asensio, A.; Kobko, N.; Dannenberg, J. J. J. Phys. Chem. A 2003, 107, 6441–6443. (81) Bandyopadhyay, D.; Bhattacharyya, D. Biopolymers 2006, 83, 313– 325. (82) Da¸bkowska, I.; Gonzalez, H. V.; Jurecˇka, P.; Hobza, P. J. Phys. Chem. A 2005, 109, 1131–1136. (83) Kelly, R. E. A.; Kantorovich, L. N. J. Phys. Chem. C 2007, 111, 3883–3892. (84) Kelly, R. E. A.; Lee, Y. J.; Kantorovich, L. N. J. Phys. Chem. B 2005, 109, 22045–22052. (85) Monajjemi, M.; Chahkandi, B. J. Mol. Struct.: THEOCHEM 2005, 714, 43–60. (86) Pacureanu, L.; Simon, Z. Int. J. Quantum Chem. 2010, 110, 1295– 1305. (87) Priyakumar, U. D.; MacKerell, A. D. Chem. ReV. 2005, 106, 489– 505. (88) Robertazzi, A.; Platts, J. A. J. Phys. Chem. A 2006, 110, 3992– 4000. (89) Roy, A.; Panigrahi, S.; Bhattacharyya, M.; Bhattacharyya, D. J. Phys. Chem. B 2008, 112, 3786–3796. (90) Sharma, P.; Mitra, A.; Sharma, S.; Singh, H. J. Chem. Sci. 2007, 119, 525–531. (91) Sˇponer, J.; Leszczynski, J.; Hobza, P. J. Mol. Struct.: THEOCHEM 2001, 573, 43–53. (92) Sˇponer, J.; Leszczynski, J.; Hobza, P. Biopolymers 2001, 61, 3– 31. (93) Calculated at the B3LYP/6-311+G(2df,p)//B3LYP/6-31G(d) level. (94) The model size affects the deformation energy for the ortho, but not para, adduct, since the favorable O-H · · · N7 hydrogen bond in the ortho adduct monomer is disrupted upon base pairing and this disruption is greater for the nucleobase than the nucleoside model. In contrast, there is no similar disruption in the most favorable geometry of the para adduct upon base pairing. (95) This is due to steric clashes between the phenoxyl group and N3H of thymine (i.e., C-H · · · H-N3 contacts). Indeed, no Hoogsteen para adduct-thymine pair is reported in the present study because optimizations result in highly distorted, unrealistic interactions between the para hydroxyl group and O4 of thymine. In contrast, the Hoogsteen natural guaninethymine pair forms a C8-H · · · O2 hydrogen bond, which occurs due to distortion between the bases to minimize repulsion between O6 (G) and O4 (T). (96) Rutledge, L. R.; Wheaton, C. A.; Wetmore, S. D. Phys. Chem. Chem. Phys. 2007, 9, 497–509.
JP105817P
